Demand Forecasting For Economic Order Quantity In .

International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013ISSN 2250-31531Demand Forecasting For Economic Order Quantity inInventory ManagementAju Mathew*,Prof.E.M.Somasekaran Nair**,Asst Prof. Jenson Joseph E****Mechanical Engineering Department, SCMS School of Engineering and Technology,Karukutty, KeralaMechanical Engineering Department, SCMS School of Engineering and Technology,,Karukutty, Kerala***Mechanical Engineering Department, SCMS School of Engineering and Technology, ,Karukutty, Kerala**Abstract-With today’s uncertain economy,companies are searching for alternative methods to keep ahead of theircompetitors.Forecasts of future demand will determine the quantities that should be purchased,produced and shipped.In this workα,two data mining methods,artificial neural network(ANN) and exponential smoothing(ES) were utilized to predict the demand of thefertilizer(Ammonium Sulphate).The training data used was the sales data of fertilizer of the previous 3 years.Demand forecasted byartificial neural network is more accurate and have less inventory costs than exponential smoothing method.Index Terms- Artificial neural network(ANN),Economic Order Quantity(EOQ), Exponential smoothing(ES),Inventory CostsI. INTRODUCTIONForecasting in general is prediction of some future event.Businesses use a variety of forecasts such as forecasts oftechnology,economy and sales of product or service.As a result,theaccuracy of demand forecasts will significantly improve theproduction schedulingcapacity planning,material requirement planning and inventory management.Although having accurate forecastshave never been easy,it has become more difficult in recent years due to increased uncertainty,complexity of business and reducedproduct life cycle.Traditionally,statistical methods such as time series analysis like exponential smoothing,weighted average,weightedmoving averages,holt’s model,winter’s model etc are used for quantitative forecasting.General problems with the time series approachinclude the inaccuracy of prediction and numerical instability.Most of the traditional timeseriesmethods are model based which aremore difficult to develop.Recently,applications of artificial neural networks have been increasing in business.One of the importantapplications of ANN is in the area of sales forecasting.Several distinguishing features of artificial neural networks make themvaluable and attractive for forecasting tasks,artificial neural networks are data driven self adaptive method.There are a few a prioriassumptions about the models for problem under study.After learning the data presented to them(a sample)ANNs can correctly inferthe unseen part of the population.II. METHODOLGYTwo methods used in this study were Artificial neural network method and Exponential smoothing method:A. Exponential smoothing methodIt calculates the smoothed series as a damping coefficient times the actual series plus 1 minus the damping coefficient times the laggedvalue of the smoothed series. The extrapolated smoothed series is a constant, equal to the last value of the smoothed series during theperiod when actual data on the underlying series are available. While the simple Moving Average method is a special case of the ES,the ES is more parsimonious in its data usage.Ft 1 α Dt (1 - α) Ftwhere: Dt is the actual valueFt is the forecasted valueα is the weighting factor, which ranges from 0 to 1t is the current time period.Notice that the smoothed value becomes the forecast for period t 1.B.ANN METHODwww.ijsrp.org

International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013ISSN 2250-31532Neural networks are computing models for informationprocessing. The most popularly used neural network modelin practice for retailsales is the feedforward multi-layer network. It is composed of several layers of basic processingunits called neurons or nodes. Beforeit can be used for forecasting, the NN modelmust be built first. Neural network model building (training)involves determining theorder of the network (the architecture)as well as the parameters (weights) of the model. NNtraining typically requires that thesampledata be splitinto a training set and a validation set. The training set isused to estimate the parameters of some candidate models,amongwhich the one that performs the best on the validationset is selected. The out-of-sample observations can beused to further test theperformance of the selected modelto simulate the real forecasting situations.Advantages Of Using Artificial Neural NetworkAdaptive learning, Self-Organisation, .Real Time Operation, and Fault Tolerance via Redundant Information Coding.III. RESEARCH ELABORATIONSProduction of Ammonium sulphate includeH2SO4 2NH3(NH3)2SO4One ton of Ammonium sulphate requires 0.6 ton of sulphur rock.Data collection:The first and foremost step of the model is collection of data.For implementation of exponential smoothing and Artificial neural networks for sale forecasting ,the monthly salesof fertilizer for last three year starting from March 2010 to march 2013 were collected.(Table.1)Forecasting of demand by Exponential Smoothing:Lt 1 α Dt 1 (1-α)Ltα 0.2Demand forecasted by Exponential smoothing is given in the table.2Total demand of Ammonium sulphate for a period from April 2013 to March 2014 165601 tonsRaw material(sulphur rock) required 0.6*165601 99360 tons.EOQ of sulphur rock 92634 tons.No of orders per year 3Ordering costs Rs 1320000Holding costs Rs8486944Inventory costs of sulphur rockForecasting of demand byArtificial neural network:MATLAB contains inbuilt NEURAL NETWORK tool box, which contains neural time series tool (ntstool). Time series tool helps inforecasting the demand of the fertilizer.Demand forecasted by ANN is given in the table.3Total demand of Ammonium sulphate for a period from April 2013 to March 2014 265217 tonsRaw material(sulphur rock) required 0.6*265217 99360 tons.EOQ of sulphur rock 33153 tons.www.ijsrp.org

International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013ISSN 2250-3153No of orders per year 8Ordering costs Rs 5280000Holding costs Rs 28179293Inventory costs of sulphur rockIV. RESULTSIn this work demand of Ammonium sulphate is forecasted by Exponential smoothing and Artificial neural network methods.EOQ andno of orders of sulphur rock is calculated with the help of forecasted demand.Number of orders per year by Exponential smoothing isthree.Number of orders per year by ANN is eight.Inventory costs,raw material holding costs and ordering costs of sulphur rock for thecoming year is obtained.A comparison is made with the help of charts.By comparing two methods a savings of 17.42% is obtained inthe case of ANN method.Inventory costs of sulphur rock for exponential smoothing ventory costs of sulphurrock for current modelFigure 1:Inventory costs of sulphur rock for exponential smoothing.Inventory costs of sulphur rock for neural network ventory costs ofsulphur rock forrecommended modelwww.ijsrp.org

International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013ISSN 2250-31534Figure 2:Inventory costs of sulphur rock for artificial neural network.V. CONCLUSIONThe current forecasting model in place at Company has brought problems due to ineffective forecasting that resulted in inaccurateinventory level. In order to help them reduce their stock outs, a forecasting model was provided along with an economic orderquantity. Finally, the economic order quantity is, optimized the order quantity for each product when an order is placed, reducing thecompanies product stock out issue. By providing and recommending the inventory control model, the results have shownimprovements in forecasting as well as in cost reduction. So, if the company follows through and implements the recommendedinventory model, they would be able to reduce the total cost by approximately 20%which is a cost reduction of for top sellingproducts. In the end, the issues the company faces would be reduced by implementing the recommended inventory model. The modelwill ensure the product is in stock, which would drive product sales and would allow the company to increase profit by forecastingaccordingly. The recommended analysis showed that simple, yet complex techniques are the key for retail success which could givethem the competitive edge.REFERENCES[1][2][3][4][5][6][7][8]Zhang G., Patuwo B. E. and Hu M. Y., Forecasting with neural networks: The state of the art, International Journal of Forecasting, 14 (1), 35-62 (1998).Atiya, F. A., & Shaheen, I. S. (1999). A comparison between neural-network forecasting techniques-case study: River flow forecasting. IEEE Transactions onNeural Networks, 10(2).Baxt, W. G. (1992). Improving the accuracy of an artificial neural network using multiple differently trained networks. Neural Computation, 4, 772–780.Chu, C.H., Widjaja, D., 1994. Neural network system for forecastings of the IEEE International Conference on Neural Networks, ing method selection. DecisionSupport Systems 12, 13–24.Borisov, A.N., Pavlov, V.A., 1995. Prediction of a continuous comparing mean square forecast errors. Journal of ForecastingGurney, K. (1997). An Introduction to Neural Networks, Routledge, ISBN 1-85728-673-1 LondonA.R Sourush, I.Nakhai Kamal Abidi,A.Behereininijad,“review on applications of artificial neural network and its future”, world applied science journal 6,2009,pp.12-18.Bo K. Wong , Thomas A. Bodnovich , Yakup Selvi, “Neural network applications in business: A review and analysis of the literature”, Decision Support Systems19 ,1997,pp. 301-320Table1.Consolidated sales data of four states.MonthConsolidated sales of four statesAMMONIUM 545www.ijsrp.org

International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013ISSN 888558Table 2. Demand forecasted by Exponential smoothingMONTHAMMONIUM 87www.ijsrp.org

International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013ISSN ble 3. Demand forecasted by Artificial neural network.FORECASTED DEMAND OF AMMONIUM SULPHATE FROM APRIL 2013 TO MARCH 2014All units are in metric tonsMONTHAMMONIUM 67Feb-148992.333333Mar-149129www.ijsrp.org

inventory level. In order to help them reduce their stock outs, a forecasting model was provided along with an economic order quantity. Finally, the economic order quantity is, optimized the order quantity for each product when an order is placed, reducing the companies product stock out issue.